We have developed and validated a series of quantitative perfusion methods including a dual bolus protocol (Christian TF et al. Radiology. 2004; 232(3):677-84). We demonstrated that all of the conclusions from the pre-clinical studies were applicable and relevant in people (Hsu L et al. J Magn Reson Imaging 2006; 23(3):315-22). One problem common to prior quantitative perfusion analyses (including our own work) has been reliance on region of interest analysis or sector-based analysis. These types of sector-based analyses are equivalent to degrading the image resolution by a factor of 10-50 and result in rather blocky representations of myocardial perfusion. Since physicians generally interpret perfusion images by watching a series of images played in a digital video loop at full resolution, we hypothesized that MRI could evaluate fully quantitative myocardial blood flow (MBF) at a pixel level based on contrast-enhanced first-pass cardiac magnetic resonance (CMR) imaging in dogs and patients. We found that myocardial blood flow could be quantified at the pixel level (32 microliters of myocardium) on CMR perfusion images and results compared well with microsphere measurements (Hsu L et al. in press). Furthermore, high-resolution pixel-wise CMR perfusion maps could detect transmural perfusion gradients. These methods may improve the objectivity of diagnosing CAD in patients.